After months of building in stealth, I'm excited to officially introduce Nen to the world.

We started with a problem I kept running into as a developer: AI agents that talked a good game but couldn't actually operate production desktop applications. Everyone was building copilots and chatbots. Nobody was building the infrastructure that lets agents operate desktop software — click through interfaces, navigate workflows, extract data, and handle the repetitive work that eats up engineering time.

You can call Claude or other models directly and get something working — that's fine for a POC or hobby project. But production means handling all the cruft before you even get into the logic of computer use: maintaining stable remote desktop connections, providing VPN proxies, keeping video recordings of automations, storing logs, building durable retries, getting structured error messages when things break. If you're an experienced engineer who thinks in terms of SLAs, availability, and error handling, you know the capability-reliability gap is huge.

So we built it ourselves.

The Problem We're Solving

Desktop enterprise software — EHRs, government portals, legacy operations tools — wasn't built to be automated. There's no API. No webhook. No export button that does what you need.

If you're building products that need to interact with these systems, you're stuck:

Legacy desktop systems with interfaces never designed for programmatic access
Workflows that span multiple applications with no common integration layer
Data locked inside systems that don't speak the same language

These aren't problems you solve with APIs or database connectors. You need agents that can operate software the way humans do — see screens, click buttons, fill out forms, adapt when interfaces change.

What We Built

Nen is the infrastructure layer for building, deploying, and scaling agents that operate desktop software.

Our agents operate software. They see screens, click buttons, navigate UIs (including legacy systems never designed for automation), and handle the repetitive work that consumes development cycles. Built for data-sensitive environments with strict guardrails and verification steps. Adaptable to your specific systems - whether Epic, Cerner, banking platforms, or proprietary tools.

We handle three things:

Infrastructure — Launching VMs, managing computer use sessions, everything you'd expect from infrastructure platforms, but for desktop computer use
Observability — Monitoring, logging, and production-grade durability.
Model flexibility — The latest computer use models in a single interface. Models are evolving fast, and we keep you on the cutting edge without the integration work

We see computer use as a component in a heterogeneous integration landscape. Use APIs when possible, database integration when it makes sense, and computer use when you need to interact with systems that weren't built for programmatic access.

We're starting with industries where legacy desktop software is the norm — healthcare, logistics, financial services — anywhere teams spend more time navigating systems than building products.

Why Now

Computer use agents have been "the next big thing" for years. The breakthrough is that foundation models can now handle visual reasoning and multi-step planning reliably enough for production use cases.

The teams that ship computer use agents first will move faster, automate what their competitors can't, and free their engineers to focus on the product instead of system access.

What's Next

We're opening our platform today because the best software hasn't been built yet.

It's going to be built by teams that deploy agents to handle desktop operations while developers focus on their core product. By organizations that automate what was previously impossible to automate. By developers who ship in weeks what used to take quarters.

Ready to see what your team can build? Book a meeting. Let's chat.